In chasemc/ghosts: Protein and small molecule analysis of bacteria by MALDI-TOF MS

SQLite

defb experiments are built upon SQLite, as included withing the RSQLite package. This provides a lightweight but powerful RDMS for handling all the data stored by defb.

What R packages does defb use to work with SQLite?

• RSQLite contains the actual RSQLite drivers and should be referenced for current information about SQLite version
• Pool "Enables the creation of object pools, which make it less computationally expensive to fetch a new object."
• DBI handles the communication with the database
• glue (handles quoting)
• not used within defb, but dplyr is also convenient to use for querying defb databases

What defb functions create the database?

The answer is pretty complicated, but some places to start looking are:

sqlTableArchitecture.R (Layouts the database structure)

and

dbCreators.R (creates a database)

How does defb handle adding new replicates to an existing experiment/database?

Let's say you have an experiment that contains a sample named "sample_1", and then later want to add a new replicate of "sample1" and merge this into the earlier experiment. This works because...

The mzML/mzXML file is stored and indexed by its hash, not by sample ID.

This hash is linked to sampleIDs in the "IndividualSpectra" table and so spectrum and peak data is always attributable to the originating mzML/mzXML.

chasemc/ghosts documentation built on May 8, 2019, 12:50 a.m.